diff --git a/litellm/tests/test_completion.py b/litellm/tests/test_completion.py
index 96ddee40a..3d8b9a1b2 100644
--- a/litellm/tests/test_completion.py
+++ b/litellm/tests/test_completion.py
@@ -26,7 +26,7 @@ def test_completion_custom_provider_model_name():
     try:
         litellm.cache = None
         response = completion(
-            model="together_ai/togethercomputer/llama-2-70b-chat",
+            model="together_ai/mistralai/Mistral-7B-Instruct-v0.1",
             messages=messages,
             logger_fn=logger_fn,
         )
@@ -287,7 +287,7 @@ def hf_test_completion_tgi():
         print(response)
     except Exception as e:
         pytest.fail(f"Error occurred: {e}")
-# hf_test_completion_tgi()
+hf_test_completion_tgi()
 
 # ################### Hugging Face Conversational models ########################
 # def hf_test_completion_conv():
@@ -1023,7 +1023,7 @@ def test_customprompt_together_ai():
         litellm.set_verbose = False
         litellm.num_retries = 0
         response = completion(
-            model="together_ai/togethercomputer/llama-2-70b-chat",
+            model="together_ai/mistralai/Mistral-7B-Instruct-v0.1",
             messages=messages, 
             roles={"system":{"pre_message":"<|im_start|>system\n", "post_message":"<|im_end|>"}, "assistant":{"pre_message":"<|im_start|>assistant\n","post_message":"<|im_end|>"}, "user":{"pre_message":"<|im_start|>user\n","post_message":"<|im_end|>"}}
         )
@@ -1599,7 +1599,7 @@ def test_completion_together_ai_stream():
     messages = [{ "content": user_message,"role": "user"}]
     try:
         response = completion(
-            model="together_ai/togethercomputer/llama-2-70b-chat", 
+            model="together_ai/mistralai/Mistral-7B-Instruct-v0.1", 
             messages=messages, stream=True, 
             max_tokens=5
         )
diff --git a/litellm/utils.py b/litellm/utils.py
index 4df06f52e..c6d855fd1 100644
--- a/litellm/utils.py
+++ b/litellm/utils.py
@@ -132,6 +132,9 @@ def map_finish_reason(finish_reason: str): # openai supports 5 stop sequences -
         return "content_filter"
     elif finish_reason == "ERROR": # openai currently doesn't support an 'error' finish reason
         return "stop"
+    # huggingface mapping https://huggingface.github.io/text-generation-inference/#/Text%20Generation%20Inference/generate_stream
+    elif finish_reason == "eos_token" or finish_reason == "stop_sequence":
+        return "stop"
     return finish_reason
 
 class FunctionCall(OpenAIObject):
diff --git a/requirements.txt b/requirements.txt
index b6a15fb7a..f3be861b8 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -3,6 +3,7 @@ litellm
 openai
 fastapi
 tomli
+pydantic>=2.5
 appdirs
 tomli_w
 backoff